Latently infected resting memory CD4+ T cells are the primary barrier to the eradication of HIV. Recently, a number of compounds have been identified that can selectively reactivate latent HIV, raising hopes that the virus can be reactivated and eliminated through immune responses, antiviral therapy, or cytopathic effects. However, the current 'gold-standard' assay for measuring the latent reservoir, the quantitative viral outgrowth assay (Q-VOA), is labor-intensive, costly, and requires cells from multiple healthy donors, making it impractical for large clinical trials. This proposal aims to improve a selective reaction monitoring-mass spectrometry (SRM- MS) assay that has been developed to measure the size of the latent HIV reservoir in patients. We are currently able to detect ~16 infected cells among a population of 80,000 uninfected CD4+ T cells, the maximum that can be loaded onto the mass spectrometer. In Aim 1, we will evaluate strategies to enrich for HIV proteins and peptides from much larger numbers of cells (up to at least 1x107) while detecting 10 or fewer infected cells. In Aim 2, we will compare the SRM-MS with traditional metrics of HIV infection including proviral DNA, cellular viral mRNA, supernatant genomic viral RNA, and the Q-VOA in cells from patients with undetectable viral loads for at least 6 months on antiretroviral therapy. The two areas to be investigated in this project are: 1. Evaluate enrichment strategies to improve the empirical sensitivity of SRM-MS. The SRM-MS assay has a theoretical sensitivity ~6-fold lower than the virus estimated to be produced by a single activated CD4+ T cell. Our primary limitation with the SRM-MS is the amount of protein that can be loaded onto the instrument: protein from approximately 80,000 CD4+ T cells. To detect infected cells in larger populations of CD4+ T cells, we will investigate HIV Gag protein and peptide enrichment strategies. These enrichment strategies have the additional advantage of improving our signal to noise ratio as non-HIV proteins are removed, further improving the sensitivity of the assay. 2. Compare the sensitivity of the SRM-MS assay to traditional measures of viral infection including proviral DNA, viral RNA transcripts and genomic RNA, and the Q-VOA using patient samples. The SRM- MS assay has considerable advantages in throughput, turnaround time, sample size, and cost and potentially also sensitivity and reproducibility compared to the Q-VOA. In this aim, we will assess whether the empirical sensitivity and reproducibility of the SRM-MS assay are sufficient for use in detecting latent HIV in patients with undetectable viral loads while on antiretroviral therapy by comparing it with traditional metrics of viral infection including provirl DNA, cellular viral mRNA, supernatant genomic viral RNA and the Q-VOA. Successful completion of these aims could provide a novel, sensitive, high-throughput, and economic assay for measuring the latent reservoir in patients enrolled in large clinical trials.